An in-vehicle active noise reduction modeling method, computer device, and storage medium based on sweep frequency identification

Abstract

The present invention provides an in-vehicle active noise reduction modeling method, a computer device and a storage medium based on sweep frequency identification. Step 1: Determine the parameters of the engine, where the parameters are the frequency range, the excitation signal time length, and the growth coefficient; Step 2: Obtain a sweep frequency signal according to the parameters in Step 1; Step 3: Calculate the error percentage according to the sweep frequency signal in Step 2; Step 4: Judge the modeling quality according to the error percentage. The present invention is used to effectively guide the problem of acoustic transfer function modeling work.

Description

Technical Field

[0001] The present invention belongs to the technical field of automotive noise reduction; specifically, it relates to an in-vehicle active noise reduction modeling method based on sweep frequency identification, a computer device, and a storage medium. Background Art

[0002] The development of passenger vehicle NVH has become increasingly perfect. Traditional NVH sound absorption and insulation technologies have a good filtering effect on high-frequency noise. However, for low-frequency noise (<300 Hz), due to its long wavelength and strong penetrability, the effect of traditional NVH means is very limited. Moreover, long-term low-frequency noise, especially the low-frequency rumbling sound inside the vehicle, will interfere with the driver's attention inside the vehicle and make people feel irritable and uncomfortable. Therefore, the impact of in-vehicle low-frequency noise on passengers inside the vehicle is very obvious, and NVH engineers are committed to controlling and even eliminating this kind of sound.

[0003] The in-vehicle active noise reduction technology is a complex system, involving many influencing factors, including hardware performance design, noise reduction microphone layout, vehicle-wide speaker frequency response performance, acoustic transfer function modeling, noise reduction algorithms, etc. Each link will have an important impact on the effect of active noise reduction.

[0004] The in-vehicle active noise reduction technology is a complex system, involving many influencing factors, including hardware performance design, noise reduction microphone layout, vehicle-wide speaker frequency response performance, acoustic transfer function modeling, noise reduction algorithms, etc. Each link will have an important impact on the effect of active noise reduction.

[0005] In the prior art, "An active noise reduction method for offline modeling of a secondary channel based on an EMFNL filter" discloses an active noise reduction method for offline modeling of a secondary channel based on a band-linear part even mirror Fourier nonlinear filter, which includes: S1: constructing the taps of the EMFNL filter and using an adaptive algorithm to identify its coefficients to achieve offline modeling of the secondary channel; S2: calculating the secondary channel estimation corresponding to the secondary channel coefficients identified in S1; S3: using the secondary channel estimation to implement the active noise reduction method.

[0006] In the prior art, "An active noise control method for synchronous modeling and control" discloses an active noise control method for synchronous modeling and control. The method includes the following steps: (1) initializing the control filter as a non-zero vector; (2) using a microphone to obtain a signal, and then converting the analog signal into a digital signal through AD conversion; (3) performing a filtering operation on the reference signal with the estimated control filter coefficients, secondary channel, and primary channel to obtain a modeling signal; (4) using an adaptive algorithm to update the control filter coefficients, secondary channel, and primary channel; (5) calculating the output signal with the updated control filter coefficients and giving it to the secondary sound source. Summary of the Invention

[0007] The present invention provides a method for modeling in-vehicle active noise reduction based on sweep frequency identification, a computer device, and a storage medium, aiming to solve the problem of effectively guiding the sound transfer function modeling work.

[0008] The present invention is implemented through the following technical solutions:

[0009] A method for modeling in-vehicle active noise reduction based on sweep frequency identification, the in-vehicle active noise reduction modeling method includes the following steps:

[0010] Step 1: Determine the parameters of the engine, and the parameters are the frequency range, the excitation signal time length, and the growth coefficient;

[0011] Step 2: Obtain the sweep frequency signal according to the parameters in Step 1;

[0012] Step 3: Calculate the error percentage according to the sweep frequency signal in Step 2;

[0013] Step 4: Judge the modeling quality according to the error percentage.

[0014] A method for modeling in-vehicle active noise reduction based on sweep frequency identification, specifically, Step 2 is:

[0015] Step 2.1: Determine the frequency range of the engine;

[0016] Step 2.2: Determine the excitation signal time length and the growth coefficient according to the frequency range in Step 2.1;

[0017] Step 2.3: Calculate the frequency variation function of the sweep frequency signal with time according to the starting frequency in Step 2.1 and the growth coefficient in Step 2.2.

[0018] A method for modeling in-vehicle active noise reduction based on sweep frequency identification, specifically, Step 3 is:

[0019] Step 3.1: Calculate the excitation signal and the error percentage according to the frequency variation function of the sweep frequency signal with time;

[0020] Step 3.2: Draw the transfer function curve according to the error percentage in Step 3.1.

[0021] A method for modeling in-vehicle active noise reduction based on sweep frequency identification, specifically, Step 2.1 is: according to the frequency response curve coverage range of the vehicle's audio system of the target vehicle model, set the starting frequency f 0 , combined with the analysis of the engine order noise characteristics of the target vehicle model, set the cut-off frequency f 1 .

[0022] A method for modeling in-vehicle active noise reduction based on sweep frequency identification, specifically, Step 2.2 is: set the time length of the excitation signal, and calculate the growth coefficient k according to the starting frequency and the ending frequency, and the growth coefficient k is:

[0023]

[0024] where t 0 is the sweep time.

[0025] A method for in-vehicle active noise reduction modeling based on sweep frequency identification. Specifically, step 2.3 is as follows: The frequency change of the sweep signal increases exponentially with time t, and the frequency change function is:

[0026] freq(t) = f 0 × k t (2)

[0027] where f 0 is the starting frequency and k is the growth coefficient.

[0028] A method for in-vehicle active noise reduction modeling based on sweep frequency identification. Specifically, step 3.1 is as follows: The output of the sweep signal drives the speaker to emit sound, and the error microphone continuously receives the audio signal. By continuously updating the adaptive filter coefficients, the convolution result of the excitation signal and the filter is calculated, and the error percentage between the convolution result and the signal received by the microphone is calculated. If it is less than 5%, the identification result is determined to be reliable; the sweep signal is a sweep sine signal, which is calculated by the following formula:

[0029] θ(t) = 2π × f 0 × (k t - 1) / log(k) (3)

[0030] y(t) = sin(2π × f 0 × (k t - 1) / log(k)) (4)

[0031] where θ(t) is the phase function and y(t) is the sweep signal function.

[0032] A method for in-vehicle active noise reduction modeling based on sweep frequency identification. Specifically, step 4 is as follows: If the error percentage meets the requirements and the transfer function curve shows a convergent trend, the acoustic transfer function modeling is completed, and the result is exported for active noise reduction program control;

[0033] If the error percentage does not meet the requirements or the transfer function curve does not converge, repeat step 2 to increase the signal time length and update formula (1).

[0034] A computer device includes a memory and a processor. The memory stores a computer program, and when the processor executes the computer program, the steps of the method described in any one of the above are implemented.

[0035] A non - transitory computer - readable storage medium stores a computer program thereon, and when the computer program is executed by a processor, the steps of the method described in any one of the above are implemented.

[0036] The beneficial effects of the present invention are:

[0037] The sound transfer function modeling time of the present invention can be freely controlled.

[0038] The sound transfer function modeling frequency range of the present invention can be adjusted according to the target requirements. BRIEF DESCRIPTION OF THE DRAWINGS

[0039] Appendix Figure 1 The flowchart of the method of the present invention.

[0040] Appendix Figure 2 The flowchart of the computer device of the present invention.

[0041] Appendix Figure 3 The schematic diagram of the identification result of the secondary channel in Example 11 of the present invention in the vehicle interior environment.

[0042] Appendix Figure 4 The schematic diagram of the noise - canceling microphone arranged at the co - driver headrest position in the present invention.

[0043] Appendix Figure 5 The experimental curve graph of the active noise - canceling function test of the present invention, where (a) is the comparison curve graph of the total sound pressure level before and after noise cancellation in the range of 30 Hz - 300 Hz, (b) is the comparison curve graph of the engine 2nd - order noise curve before and after noise cancellation, (c) is the comparison curve graph of the engine 4th - order noise curve before and after noise cancellation, and (d) is the comparison curve graph of the engine 6th - order noise curve before and after noise cancellation. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0044] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all of the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without making creative efforts belong to the scope of protection of the present invention.

[0045] Embodiment 1

[0046] An in - vehicle active noise - canceling modeling method based on sweep - frequency identification, the in - vehicle active noise - canceling modeling method includes the following steps:

[0047] Step 1: Determine the parameters of the engine, and the parameters are the frequency range, the excitation signal time length, and the growth coefficient;

[0048] Step 2: Obtain a sweep - frequency signal according to the parameters in Step 1;

[0049] Step 3: Calculate the error percentage based on the swept-frequency signal in Step 2;

[0050] Step 4: Judge the modeling quality according to the error percentage.

[0051] Embodiment 2

[0052] The difference between Embodiment 2 and Embodiment 1 of this application lies only in:

[0053] A method for modeling active noise reduction in a vehicle based on swept-frequency identification, specifically, Step 2 is as follows,

[0054] Step 2.1: Determine the frequency range of the engine;

[0055] Step 2.2: Determine the excitation signal time length and growth coefficient according to the frequency range in Step 2.1;

[0056] Step 2.3: Calculate the frequency variation function of the swept-frequency signal with time according to the starting frequency in Step 2.1 and the growth coefficient in Step 2.2.

[0057] Embodiment 3

[0058] The difference between Embodiment 3 and Embodiment 2 of this application lies only in:

[0059] A method for modeling active noise reduction in a vehicle based on swept-frequency identification, specifically, Step 3 is as follows,

[0060] Step 3.1: Calculate the excitation signal and the error percentage according to the frequency variation function of the swept-frequency signal with time;

[0061] Step 3.2: Draw the transfer function curve according to the error percentage in Step 3.1.

[0062] Embodiment 4

[0063] The difference between Embodiment 4 and Embodiment 3 of this application lies only in:

[0064] A method for modeling active noise reduction in a vehicle based on swept-frequency identification, specifically, Step 2.1 is as follows, set the starting frequency f according to the frequency response curve coverage range of the vehicle's audio system of the target vehicle model 0 , combined with the analysis of the engine order noise characteristics of the target vehicle model, set the cut-off frequency f 1 .

[0065] Embodiment 5

[0066] The difference between Embodiment 5 and Embodiment 4 of this application lies only in:

[0067] An in-vehicle active noise reduction modeling method based on sweep frequency identification. Specifically, in step 2.2, set the time length of the excitation signal, and calculate the growth coefficient k according to the starting frequency and the ending frequency. The growth coefficient k is:

[0068]

[0069] where t 0 is the sweep time.

[0070] It can set the frequency range of active noise reduction in combination with the vehicle's overall frequency response characteristics of the target vehicle model, and can provide the target noise reduction frequency range for generating the excitation signal.

[0071] Example Six

[0072] The difference between the sixth embodiment and the fifth embodiment of this application is only that:

[0073] An in-vehicle active noise reduction modeling method based on sweep frequency identification. Specifically, in step 2.3, the frequency change of the sweep signal increases linearly with time t, and the frequency change function is:

[0074] freq(t) = f 0 ×k t (2)

[0075] where f 0 is the starting frequency and k is the growth coefficient.

[0076] According to the parameters output in 2.1, set the signal time length to be usually between 10 s and 20 s; the frequency growth coefficient can be calculated through the frequency range, the excitation signal time length and the growth coefficient.

[0077] Calculate the functional relationship between the frequency and time according to the starting frequency and the growth coefficient.

[0078] Example Seven

[0079] The difference between the seventh embodiment and the sixth embodiment of this application is only that:

[0080] An in-vehicle active noise reduction modeling method based on sweep frequency identification. Specifically, in step 3.1, the output of the sweep signal drives the speaker to emit sound, and the error microphone receives the audio signal in real time. By continuously updating the adaptive filter coefficients, calculate the convolution result of the excitation signal and the filter, and calculate the error percentage between the convolution result and the signal received by the microphone. If it is less than 5%, it is determined that the identification result is reliable; the sweep signal is a sweep sine signal, and is calculated by the following formula:

[0081] θ(t) = 2π × f 0 ×(k t -1) / log(k) (3)

[0082] y(t) = sin(2π × f 0 × (k t - 1) / log(k)) (4)

[0083] Where θ(t) is the phase function and y(t) is the swept - frequency signal function.

[0084] According to the functional relationship, calculate the swept - frequency excitation signal and drive the speaker to produce sound, and the acoustic transfer function can be iterated according to the LMS algorithm to complete the swept - frequency process; and if the error between the convolution result of the FIR filter and the original noise is less than 5%, it is confirmed that the acoustic transfer function modeling process converges effectively.

[0085] Example Eight

[0086] The difference between the eighth embodiment of this application and the seventh embodiment is only that:

[0087] A method for in - vehicle active noise reduction modeling based on swept - frequency identification. Specifically, in step 4, if the error percentage meets the requirements and the transfer function curve shows a converging trend, the acoustic transfer function modeling is completed, and the result is exported for active noise reduction program control;

[0088] If the error percentage does not meet the requirements or the transfer function curve does not converge, repeat step 2 to increase the signal time length and update formula (1); the error percentage and the convergence of the transfer function curve are determined by the tester according to the actual situation.

[0089] Step 4 is to draw the acoustic transfer function curve, and the tester judges the convergence state of the transfer function curve according to experience. If the tester judges that the convergence is unstable, return to step 2 to appropriately increase the time length, and repeat the content from step 2 to step 4.

[0090] The tester determines that the convergence is stable according to the objective error percentage, and the exported transfer function result can provide accurate input for active noise reduction, which is beneficial to ensuring the stability of active noise reduction performance.

[0091] Example Nine

[0092] The difference between the ninth embodiment of this application and the eighth embodiment is only that:

[0093] A computer device includes a memory and a processor. The memory stores a computer program, and when the processor executes the computer program, it implements the steps of the method described in any one of the above.

[0094] Example Ten

[0095] The difference between the tenth embodiment of this application and the ninth embodiment is only that:

[0096] A non-transitory computer-readable storage medium stores a computer program thereon, and when the computer program is executed by a processor, the steps of the method described in any one of the above are implemented.

[0097] Example XI

[0098] The difference between Example XI and Example X of this application is only that:

[0099] Using the swept-frequency identification method combined with a specific vehicle model, the acoustic transfer function modeling results are as follows:

[0100] Using the above-derived swept-frequency formula and combining with the frequency response characteristics of the speaker, the swept-frequency range is set from 50 Hz to 300 Hz during the research process, the time is 10 s, and μ is the iteration step size, which is set to 0.2. The identification result of the secondary channel in the vehicle interior environment is as Figure 3 shown.

[0101] Real vehicle layout and test results

[0102] The vehicle is a test vehicle. Compared with the production vehicle, the test vehicle has more noise, which is more beneficial to the effect of active noise reduction. However, on the production vehicle with less noise, the effect of active noise reduction will be reduced.

[0103] The noise reduction microphone is arranged at the position of the co-driver's headrest, the controller is arranged in the trunk, and the right front door bass speaker is driven to emit sound. The noise reduction microphone is arranged at the co-driver's headrest. The standard microphone used for testing is arranged at the same position as the noise reduction microphone, and the rest of the test equipment and test personnel are in the back row. Connect the computer and the controller through the emulator. First, brush the compiled acoustic channel identification program and the noise reduction control program into the controller to achieve secondary channel identification and noise reduction control, as Figure 4 shown.

[0104] Conduct an active noise reduction function test in the chassis dynamometer laboratory. Obtain the coefficients of the secondary channel by the swept-frequency method, and import the coefficients into the noise reduction program for active noise reduction test. The gear is locked in the second gear, and the vehicle accelerates gently for 20 seconds, and the engine speed rises from 1000 r / min to 4000 r / min. By collecting and analyzing the noise data through the standard microphone, it is found that the noise reduction effect of the second, fourth, and sixth orders of the engine within 300 Hz is obvious, and the noise of the 2nd, 4th, and 6th orders is reduced by about 10 dB, and the maximum reduction of the OA value from 30 Hz to 300 Hz is 5 dB. The test results are as Figure 5 shown.

Claims

1. An in-vehicle active noise reduction modeling method based on sweep frequency identification, characterized in that, the in-vehicle active noise reduction modeling method comprises the following steps: Step 1: Determine the parameters of the engine, and the parameters are frequency range, excitation signal time length, and growth coefficient; Step 2: Obtain a sweep frequency signal according to the parameters in Step 1; Step 3: Calculate the error percentage according to the sweep frequency signal in Step 2; Step 4: Judge the modeling quality according to the error percentage; The specific content of Step 2 is, Step 2.1: Determine the frequency range of the engine; Step 2.2: Determine the excitation signal time length and growth coefficient according to the frequency range in Step 2.1; Step 2.3: Calculate the frequency change function of the sweep frequency signal with time according to the starting frequency in Step 2.1 and the growth coefficient in Step 2.2; Specifically, step 2.2 is to set the time length of the excitation signal and calculate the growth coefficient k according to the starting frequency f 0 and the cut-off frequency f 1 , and the growth coefficient k is as follows: Among them, t 0 is the sweep time; it can set the frequency range of active noise reduction in combination with the vehicle frequency response characteristics of the target vehicle model, and can provide the target noise reduction frequency range for generating the excitation signal; The specific content of Step 2.3 is that the frequency change of the sweep frequency signal increases exponentially with time t, and the frequency change function is: f(t) = f 0 × k t (2) where f 0 is the starting frequency, k is the growth coefficient, and f 1 is the cut-off frequency; According to the parameters output in 2.1, set the signal time length to be usually between 10s and 20s; the frequency growth coefficient can be calculated through the frequency range, excitation signal time length, and growth coefficient; The specific content of Step 3 is, Step 3.1: Calculate the excitation signal and error percentage according to the frequency change function of the sweep frequency signal with time; The specific content of Step 3.1 is that the output of the sweep frequency signal drives the speaker to emit sound, the error microphone receives the audio signal in real time, by continuously updating the adaptive filter coefficients, calculate the convolution result of the excitation signal and the filter, calculate the error percentage between the convolution result and the signal received by the microphone, and if it is less than 5%, it is determined that the identification result is reliable; the sweep frequency signal is a sweep frequency sine signal, and is calculated by the following formula: θ(t) = 2π × f 0 × (k t - 1) / log(k) (3) y(t) = sin(2π×f 0 ×(k t - 1) / log(k)) (4) where, θ(t) is the phase function, and y(t) is the sweep frequency signal function; according to the functional relationship, calculate the sweep frequency excitation signal and drive the speaker to emit sound, and the acoustic transfer function can be iterated according to the LMS algorithm to complete the sweep frequency process; and if the error between the convolution result of the FIR filter and the original noise is less than 5%, it is confirmed that the acoustic transfer function modeling process converges effectively; Step 3.2: Draw a transfer function curve according to the error percentage in Step 3.1; The specific content of Step 4 is that if the error percentage meets the requirements and the transfer function curve shows a converging trend, the acoustic transfer function modeling is completed, and the result is exported for active noise reduction program control; If the error percentage does not meet the requirements or the transfer function curve does not converge, repeat Step 2 to increase the signal time length and update formula (1); the error percentage and the convergence of the transfer function curve are determined by the tester according to the actual situation; Step 4 is to draw an acoustic transfer function curve, and the tester judges the convergence state of the transfer function curve according to experience. If the tester judges that the convergence is unstable, return to Step 2 to appropriately increase the time length, and repeat the content of Steps 2 to 4.

2. The in-vehicle active noise reduction modeling method based on sweep frequency identification according to claim 1, characterized in that, Specifically, step 2.1 is to set the starting frequency f according to the frequency response curve coverage range of the vehicle audio system of the target vehicle model 0 , and in combination with the analysis of the engine order noise characteristics of the target vehicle model, set the cut-off frequency f 1 .

3. A computer device, comprising a memory and a processor, the memory stores a computer program, characterized in that, when the processor executes the computer program, it implements the steps of the method described in any one of claims 1-2.

4. A non-transitory computer-readable storage medium having a computer program stored thereon, wherein, when the computer program is executed by a processor, the steps of the method according to any one of claims 1-2 are implemented.

Images